Large-scale printing systems, such as rotogravure printing presses, feed a continuous web of material, typically paper, through printing machinery that forces the web into contact with one or more rotogravure printing cylinders which, in turn, print images onto the web in a standard manner. Thereafter, the web is cut into individual pages or signatures which are collated to produce, for example, newspapers, books, magazines, etc. A common and recurring problem in large-scale printing systems is the occurrence of web breaks, which happen when the web tears while the web is being fed through the individual components of the printing system. Upon the occurrence of a web break, the printing system must be shut down, the torn web must be dislodged from the individual components of the printing system and then the web must be re-fed through the printing system, all of which takes a considerable amount of time and results in wasted paper and ink. Furthermore, in some instances, web breaks may result in damage to components of the printing system.
While web breaks are a common problem in the printing industry, the reasons or conditions that lead to the occurrence of any particular web break vary a great deal. In fact, web breaks may be caused by different factors or by different combinations of factors at different times in the same printing system. Generally, web breaks are avoided by having a local expert, such as a printing press foreman, oversee the press conditions and make suggestions for changes based mainly on past experiences with web breaks, trial and error and general rules of thumb. While some of these approaches are successful in decreasing the incidence of web breaks in the short term, web breaks usually reappear later with very little indication as to the real cause of the reappearance. Furthermore, while local printing experts are usually capable of determining the general cause of any particular web break after the web break has occurred and, moreover, are generally capable of altering press conditions to eliminate a particular cause of a web break in the short term, there is no guarantee that the altered conditions will not result in further web breaks for other reasons or that the press conditions suggested by the local expert will be implemented in the press for a long period of time.
It is generally known that one of the most common conditions leading to the occurrence of a web break is excessive tension within the web at one or more locations within the printing system. Generally speaking, a discrete amount of tension must be present PATENT in the web to assure that the different printing cylinders of the printing system begin to register on the web at the same location. Slack within the web may cause misalignment between the different images which, in turn, produces an inferior product. However, as noted above, too much tension at any particular location leads to web breaks. Unfortunately, even with this rule of thumb, it is not generally known what the tension at any particular web location should be to decrease the likelihood of web breaks or, for that matter, why some tensions are better than others.
Recently, it has been suggested to use an expert system to determine the causes of problems, such as web breaks, within a printing press. In particular, the above-identified parent application on which this application relies for priority, is directed to the use of a decision-tree induction analysis that identifies conditions leading to a particular result, such as web breaks, within a printing system. In general, expert systems are used to mimic the tasks of an expert within a particular field of knowledge or domain, or to generate a set of rules applicable within the domain. In these applications, expert systems must operate on objects associated with the domain, which may be physical entities, processes or even abstract ideas. Objects are defined by a set of attributes or features, the values of which uniquely characterize the object. Object attributes may be discrete or continuous.
Typically, each object within a domain also belongs to or is associated with one of a number of mutually exclusive classes having particular importance within the context of the domain. Expert systems that classify objects from the values of the attributes for those objects must either develop or be provided with a set of classification rules that guide the system in the classification task. Some expert systems use classification rules that are directly ascertained from a domain expert. These systems require a "knowledge engineer" to interact directly with a domain expert in an attempt to extract rules used by the expert in the performance of his or her classification task.
Unfortunately, this technique usually requires a lengthy interview process that can span many man-hours of the expert's time. Furthermore, experts are not generally good at articulating classification rules, that is, expressing knowledge at the right level of abstraction and degree of precision, organizing knowledge and ensuring the consistency and completeness of the expressed knowledge. As a result, the rules that are identified may be incomplete while important rules may be overlooked. Still further, this technique assumes that an expert actually exists in the particular field of interest. Even if an expert does exist, the expert is usually one of a few and is, therefore, in high demand. As a result, the expert's time and, consequently, the rule extraction process can be quite expensive.
It is known to use artificial intelligence within expert systems for the purpose of generating classification rules applicable to a domain. For example, an article by Bruce W. Porter et al., Concept Learning and Heuristic Classification in Weak-Theory Domains, 45 Artificial Intelligence 229-263 (1990), describes an exemplar-based expert system for use in medical diagnosis which removes the knowledge engineer from the rule extraction process and, in effect, interviews the expert directly to determine relevant classification rules.
In this system, training examples (data sets that include values for each of a plurality of attributes generally relevant to medical diagnosis) are presented to the system for classification within one of a predetermined number of classes. The system compares a training example with one or more exemplars stored for each of the classes and uses a set of classification rules developed by the system to determine the class to which the training example most likely belongs. A domain expert, such as a doctor, either verifies the classification choice or instructs the system that the chosen classification is incorrect. In the latter case, the expert identifies the correct classification choice and the relevant attributes, or values thereof, that distinguish the training example from the class initially chosen by the system. The system builds the classification rules from this information, or, if no rules can be identified, stores the misclassified training example as an exemplar of the correct class. This process is repeated for training examples until the system is capable of correctly classifying a predetermined percentage of new examples using the stored exemplars and the developed classification rules.
A patent to Karis (U.S. Pat. No. 5,521,844) discloses a case-based expert system that may be used to aid in the identification of the cause of a particular problem, such as a web break, in a printing system. The expert system disclosed in the Karis patent stores data related to a set of previous printing runs or cases in which the problem, e.g., a web break, actually occurred. An expert then goes through the cases and identifies the most likely reason or reasons that the problem occurred in each case. These reasons are then stored in the memory of the expert system and, thereafter, the stored cases, along with the cause and effect reasoning provided by the expert are used to classify the cause(s) of the problem when it arises later. Unfortunately, the Karis system requires the use of an expert to originally identify the most probable cause(s) of the problem and, thus, is totally dependent on the expert's knowledge and reasoning. The Karis system does not identify causes which were never identified by the expert because, for example, the expert did not connect the problem to a particular cause or because the cause did not result in the problem in one of the cases reviewed by the expert. Furthermore, the Karis system does not store or collect data pertaining to printing runs in which the problem did not occur. As a result, the Karis system cannot perform a data mining technique, i.e., one in which causes are determined based on the data from both printing runs in which the problem did occur and printing runs in which the problem did not occur.
Other artificial intelligence methods that have been used in expert systems rely on machine induction in which a set of induction rules are developed or are induced directly from a set of records, each of which includes values for a number of attributes of an object and an indication of the class of the object. An expert then reviews the induced rules to identify which rules are most useful or applicable to the classification task being performed. Such a system is disclosed in the above-identified parent application. This method has the advantage of using the expert in a way that the expert is accustomed to working, that is, identifying whether particular rules are relevant or useful in the classification task. It should be noted, however, that all of the relevant attributes of the objects being classified must be identified and data for those attributes must be provided within the records in order for the system to induce accurate and complete classification rules.
A classic example of a pure machine induction technique is described in an article by J. R. Quinlan, Induction of Decision Trees, 1 Machine Learning 81-106 (1986), the disclosure of which is hereby incorporated by reference herein. This technique searches through relations between combinations of attribute values and classes of objects to build an induction tree which is then used to generate precise classification rules. During operation, the Quinlan method calculates a statistical measurement, referred to as an information gain value, for each of a set of attributes and chooses the attribute with the highest information gain value at a root of the tree. The attribute values associated with chosen attribute are then identified as nodes of the tree and are examined. If all of the data records associated with a node are all of the same class, the node is labeled as a leaf or endpoint of the induction tree. Otherwise, the node is labeled as a branching point of the induction tree. The method then chooses a branching point, calculates the information gain value for each of the remaining attributes based on the data from the records associated with the chosen branching point, chooses the attribute with the highest information gain value and identifies the attribute values of the chosen attribute as nodes which are examined for leaves and branching points. This process is repeated until only leaves remain within the induction tree or until, at any existing branching point, there are no attributes remaining upon which to branch. After an induction tree is constructed, classification rules are generated therefrom by tracing a path from a particular leaf of the induction tree to the root of the induction tree or vice versa.
As noted above, choosing the appropriate variables or attributes for such an expert system is an important step in identifying the cause of a problem such as web breaks. Without the appropriate choice of attributes, the expert system can be practically useless in actually determining the causes of problems such as web breaks in a printing system.